Cognitive Dissonance: Reexamining a Pivotal Theory in Psychology - Eddie Harmon-Jones 2019
In Search of the Motivation for Dissonance Reduction: The Drive to Lessen Aversive Consequences
The Role of The Self in Dissonance
Joel Cooper
More than 60 years ago, Leon Festinger (1957) made the marvelously elegant suggestion that inconsistency between pairs of cognitive elements causes the psychological discomfort known as cognitive dissonance. That statement, and a modest number of supporting postulates, spawned more than 2,000 empirical investigations examining the nonobvious predictions of the theory and expanding its reach into new areas of applicability.
The longevity of dissonance at or near the center stage of social psychology is impressive. The early days of dissonance research were characterized by novel, high-impact experimentation conducted by clever and intuitive experimenters. But this alone could not have been sufficient to carry the dissonance tradition through six decades. More impressive than the impact of the original experiments was Festinger’s foresight to recognize the careful interplay between motivation and cognition. Festinger recognized, before it was fashionable, that knowledge of the environment and knowledge of one’s own behaviors, attitudes, and emotions were represented cognitively and that it was the relationship among the cognitive representations that prompted motivation. Thus, when social psychology strayed toward the purely motivational, dissonance theory was present to remind the field of the importance of cognitive representations of the person and the environment. Similarly, and perhaps more important, when social psychology strayed toward the purely cognitive, dissonance theory reminded us that cognitive representations lead to motivation and that motivation affects the representations themselves (cf. Kunda, 1990, 1999). Thus, in addition to being a provocative theory in its own right, dissonance’s reliance on the cognitive and the motivational provided an important middle ground to which social psychology could consistently return.
In this chapter, I examine the motivational basis of dissonance reduction. Why does inconsistency motivate people to go about the difficult task of cognitive change? What makes people experience arousal and negative affect? For Festinger, the answer lay in the mere presence of inconsistency. But more recent work has suggested alternative mechanisms for change, including the elimination of aversive consequences and the bolstering of the self-concept. In the current chapter, I consider the difference between motivational views that focus on the self and those that focus on behavioral consequences. Although there is considerable wisdom drawn from the representation of the self, I argue that the motivation for dissonance reduction arises from the perception of aversive consequences and that changes of attitudes that generally follow from dissonance arousal are at the service of rendering those consequences nonaversive.
CURING THE “BUT ONLYS”: FROM INCONSISTENCY TO AVERSIVE OUTCOMES
When dissonance theory was merely 25 years old, Russell Fazio and I set out to put together much of the research that had accumulated over dissonance’s first quarter century. We wanted to see if comprehensive sense could be made of a theory that had an impressive history but that was saddled with a bad case of the “but onlys.” Dissonance arose from inconsistent cognitions, but only if there was free choice to act counterattitudinally; but only if there was commitment to the counterattitudinal act; but only if an unwanted consequence had occurred; but only if the consequence was foreseeable, and so forth.
Fazio and I wanted to see if we could plot the best fitting theoretical curve that passed through the numerous replications of dissonance and the impressive list of “but only” exceptions. In the end, we concluded that the best fitting curve did not pass through cognitive inconsistency at all (Cooper & Fazio, 1984). Rather, we concluded that dissonance was a state of arousal caused by behaving in such a way as to feel personally responsible for bringing about an aversive event. If that arousal was not misattributed to another source or reduced in some other way, dissonance arousal became dissonance motivation, an instigation to engage in cognitive changes for the fundamental purpose of rendering the consequence nonaversive. That is why the consequence, the unwanted result of a person’s behavior, is so critical in driving the dissonance engine. It is the essential ingredient, the primary reason for people engaging in the effort of dissonance reduction.
The new look model was intended not as a new theory, but rather as a change in emphasis of a venerable theory. We suggested that the emphasis on inconsistency in causing dissonance arousal had been misplaced. Because acting inconsistently usually has the potential to result in an aversive or unwanted consequence, inconsistency is a reasonable stand-in variable for aversive consequences in producing dissonance. However, it is not the inconsistency per se that causes the arousal, but rather the result of that inconsistency, the unwanted consequence. When viewed this way, the “but onlys” fall into place, not as exceptions to the theory, but as explicable and important components of dissonance.
Nonetheless, the difference in the two approaches is not trivial. On those occasions in which inconsistency can be separated from aversive consequences, it is the latter and not the former that holds up to empirical test (Cooper, 2007). Empirically, Cooper and Worchel (1970) found that inconsistency was not a sufficient condition for dissonance arousal if aversive consequences were lacking. Scher and Cooper (1989) showed that inconsistency was neither sufficient nor necessary for dissonance arousal if the production of aversive consequences had been carefully disentangled from the cognitive inconsistency.
WHAT IS AN AVERSIVE CONSEQUENCE?
Meeting the Positivistic Test
In our new look paper (1984), Fazio and I defined aversive consequences positivistically rather than a priori. According to the positivistic view, the ultimate authority for theoretical terms is how they are experienced by the perceiver. Similarly, in our view, aversive consequences are best defined by the way they are experienced by the actor rather than by an abstract definition in advance of an act that fails to consider the actor’s experience. We defined an aversive consequence as the real or potential result of behavior that a person would rather have not brought about. The emphasis in this definition is on the actor who determines, by his or her experience of an event, whether it is wanted or unwanted. What makes a consequence aversive is the way an outcome of an act is perceived by the actor. If the actor perceives the consequence as unwanted, then it successfully meets the test of being an aversive consequence.
The experiment by Cooper and Worchel (1970) established a procedure that perhaps became the paradigmatic example of an aversive consequence. Worchel and I had student volunteers participate in a procedure substantially similar to the original well-known induced-compliance experiment of Festinger and Carlsmith (1959). In the original study, participants engaged in a truly boring task and then were induced, for either a low or high monetary inducement, to tell a waiting participant (actually, a confederate of the experimenter) that the task was truly exciting and interesting. Festinger and Carlsmith had reasoned that the inconsistency between saying the task was interesting while privately believing that the task was deadly dull was sufficient to cause dissonance. As is well known, they also established and confirmed the nonobvious prediction that acting inconsistently for a small monetary inducement would create more dissonance and thus lead to more attitude change than acting inconsistently for a large inducement. People came to believe that the task was more interesting if they had lied to the waiting participant for only a meager amount of financial inducement.
Worchel and I argued that it was not the inconsistency between saying the task was interesting and believing it was dull that led to dissonance, but rather the unwanted event of having duped a fellow student to look forward to an exciting experience. The participant knew full well that the waiting participant was in for an immense letdown. Indeed, we found that if the student who was waiting explicitly stated that he believed the participant and was looking forward to performing the task, there was dissonance-produced attitude change. But if that consequence was removed by the waiting confederate, indicating that he did not believe the participant, then, despite the participant’s having behaved inconsistently, his actions produced no dissonance.
From that study, and several others that confirmed its basic tenet (e.g., Blackman, Keller, & Cooper, 2016; Cooper, Zanna, & Goethals, 1974; Hoyt, Henley, & Collins, 1972; Norton, Monin, Cooper, & Hogg, 2003), it might appear that an aversive consequence is necessarily characterized as doing something bad to someone else, like duping a fellow student, or bringing about an unwanted policy which could harm hundreds of students. Undoubtedly, these are unwanted consequences for most people and serve to initiate the arousal of dissonance. However, causing negative consequences to others and bringing about unwanted political ends are but two of the myriad behavioral outcomes that a person may find aversive. Rejecting a particular consumer item in the research procedure made famous by Brehm (1956) and his colleagues causes the potentially aversive consequence to oneself of giving up the positive benefits associated with that choice and accepting all of the negative features associated with the chosen alternative. Indeed, people may differ in their opinions of the kinds of consequences they find aversive. My list may bear significant similarities to yours, although differences would almost certainly exist. I may think it is particularly aversive to dupe a fellow student, but someone else may find it pleasant to have successfully completed a task requested by the experimenter.
Normative and Ideographic Standards
The fact that people may differ in what they find aversive does not plunge us toward definitional chaos. Regularities exist. Some behavioral outcomes violate generally shared expectations of what is appropriate in a particular situation. That is, they violate normative standards of behavior. Most people in a culture know that it is unacceptable to bring harm to others or knowingly and freely create harmful policies that affect one’s university, society, or social group. As Stone and Cooper (2001, 2003) have argued, the discrepancy between one’s behavioral outcome and the normative standard gives that behavior its psychological meaning. For most people, violating normative standards designates the consequence as aversive and unwanted and drives the arousal of dissonance.
On the other hand, behavior can also be discrepant from ideographic, or personal standards, and the violation of personal standards also can create dissonance. Years ago, Cooper and Scalise (1974) showed that introverts, as measured by the Myers-Briggs Type Indicator, experience dissonance from different behavioral outcomes than do extraverts. To the extent that personal, ideographic standards are made salient, behavioral outcomes that violate those standards serve as unwanted consequences. The standard that is made accessible in the situation gives meaning to a particular behavioral outcome. If normative standards are most easily accessible—and they usually are—then behavioral outcomes that violate those standards will cause dissonance. If personal standards become accessible, then behaviors that violate such standards will be considered aversive and lead to the arousal of dissonance. Stone and Cooper (2003) showed that making personal standards salient by having participants engage in a scrambled sentence task facilitated their use of personal standards in considering whether a consequence was aversive.
In essence, then, a behavioral consequence is aversive only when it is psychologically measured against a standard. Typically, that standard is defined by cultural and societal norms, meaning that it is consistent with what most people find aversive about a particular outcome. Occasionally, through chronic accessibility of a personal standard or through an environmental prime, a personal standard becomes highly accessible and then that standard functions as the measuring stick for determining aversiveness (Stone & Cooper, 2001).
THE ROLE OF THE SELF IN DISSONANCE
What systematic role does the self play in the dissonance process? There have been several fascinating and provocative theoretical accounts that have implicated self-esteem in the dissonance process. For E. Aronson (1968, 1992), a person’s perception of him- or herself as a competent and moral person necessarily forms one of the cognitions that lead to dissonance arousal. The inconsistent cognition arises from the representation of behavior that compromises or threatens the person’s preferred assessment of being a highly moral and competent individual. For Steele and his colleagues (e.g., Steele, 1988; Steele & Liu, 1983), dissonance reduction is a strategy designed to protect a person’s global feeling of self-worth and self-esteem.
Counterattitudinal behavior threatens the self and thus needs to be dealt with, often through attitude change. Ironically, these two versions of the role of self-esteem have made dramatically different predictions. E. Aronson’s self-consistency view argues that the typical dissonance situation is arousing only for people with high self-esteem, whereas self-affirmation theory argues that people with low self-esteem are primarily upset or threatened by attitude- inconsistent behavior (Steele, Spencer, & Lynch, 1993). These predictions stem from fundamentally different views of the self. Self-consistency views the self as an expectancy, with high self-esteem serving as a more exacting and stringent expectation for one’s behavior. For self-affirmation, high self-esteem is a resource that can be brought to bear to buffer the actor against the experience of discomfort.
In our self-standards model of dissonance, Jeff Stone and I took the position that there is no consistent or systematic role for the self in the dissonance process (Stone & Cooper, 2001). Dissonance is not limited to people of high self-esteem, as suggested by self-consistency, nor is it limited to people with low self-esteem, as indicated by self-affirmation. It is not a process that occurs only for introverts, high achievers, or people high in nurturance needs, and so forth. To the contrary, dissonance is a process that is aroused in all people whenever they are responsible for bringing about an aversive consequence. As I suggested previously, the self may become involved in helping to determine when a behavioral consequence is considered aversive. If a personal, self-relevant standard is promoted to a position of high accessibility in a person’s cognitive field, then behavioral outcomes that are at variance with the self-standard will create dissonance. But this is a very different view from the totalitarian sense of self in which one aspect of self (e.g., self-esteem) dominates and determines the existence of dissonance.
A CASE IN POINT: IS DISSONANCE SELF-AFFIRMATION?
The general proposition of self-affirmation is that people are concerned with the integrity of their self-worth and will take great pains to protect it. There is nothing in dissonance theory that contradicts this important message. The question at issue is whether dissonance reduction may be conceived solely as a strategic maneuver to accomplish self-affirmation. What Steele and his colleagues have contended is that participants’ changes of the representations of their attitudes in dissonance experiments are merely strategic maneuvers to satisfy the self-affirmation motive. The theory holds that people who commit counterattitudinal acts see that they have compromised their integrity. There is nothing special about a dissonant act; it is simply one more way that people can have their self-integrity threatened.
In self-affirmation’s view, people with shaky self-concepts (i.e., people with low self-esteem) are the ones most threatened. They will be most in need of finding a way to restore their self-integrity. However, in dissonance experiments, there is only one way provided for them: They can change their attitudes, so that their behavior no longer seems like a compromise to their integrity. Coming to believe what they said is a way for participants in a dissonance experiment to restore their self-integrity and to accomplish the task of self-affirmation. There are other ways that the self can be affirmed if people are given the opportunity. If, after performing some dastardly, consequential, counterattitudinal act, one can think of positive features that support her or his self-integrity, then, according to the self-affirmation approach, consequential, counterattitudinal behavior will not lead to attitude change.
Thus, Steele and Liu (1983) found that if people are reminded of their basic and important values by, for example, filling out an Allport—Vernon—Lindzey Study of Values (AVL; Allport, Vernon, & Lindzey, 1960) then counterattitudinal behavior does not lead to dissonance-produced attitude change.
If it were true that merely thinking about one’s good attributes eliminates cognitive dissonance, it would be a major challenge to the new look model. This is not just because it was unpredicted, but because it is at variance with the presumed motivational underpinning of dissonance. Why do people change their attitudes after counterattitudinal advocacy or after making difficult decisions among choice alternatives? As mentioned previously, it is to render the consequence of the behavior nonaversive. If you convince fellow voters to elect a disliked candidate to your city council, for example, thinking about how good a tennis player you are hardly solves the dilemma in which you placed yourself—assuming the dilemma is that the wrong councilperson has been elected. Only changing your attitude to favor the newly elected councilperson will render the outcome nonaversive. Of course, that is the crux of the theoretical difference: Is dissonance reduction a way to deal with the arousal created by the behavioral outcome, or is it merely a strategy of self-affirmation?
Consider the following thought-experiment: You are a person who thinks it is important to provide services for the handicapped. Nonetheless, an experimenter convinces you to write an essay advocating a cut in handicapped services at your university. You feel upset and negatively aroused at what you have done. If the motivation posited by self-affirmation is correct, you have all the wherewithal at your disposal to reduce your discomfort and reaffirm yourself any time you wish to do it. All of the ingredients you need are currently in your head. There are values you possess, abilities you have, accomplishments you have produced—all of these and more are waiting for you in memory if you only choose to use them. Any of those mnemonic cognitions can be used for self-affirmation and should be every bit as good as, for example, the filling out of an AVL, which forms the self-affirmation manipulation in Steele and Liu’s (1983) experiment. The question is, will you use these memory traces? Will you end your discomfort by conjuring up the recollection of some of your positive values or achievements?
Most probably, this would not happen. The results of so many experiments in which people advocated attitude-discrepant positions argue that people will not make use of the potentially self-affirming cognitions they already possess in memory. If they did—and surely such cognitions were available for use—then counterattitudinal advocacy would never lead to attitude change. It would seem so much more efficient and require so little effort to think quickly of your prowess at tennis or of the great meal you cooked the other night. Even these thoughts are only indirect ways of restoring the self-concept that was threatened by your heartless, noncompassionate essay. It would seem so much more effective to take a frontal attack on your seeming lack of concern for the handicapped by remembering how compassionate you typically have been toward people in need. There must have been some occasion in which your compassion and concern come to mind. However, we know that people do not conjure up such cognitions because cognitions that could provide such direct self-affirming information are always available in memory but are not used by people in any of the research in counterattitudinal advocacy.
It could be argued that, although they are potentially available, such cognitions are largely inaccessible in the context of a psychology experiment. You chose to make a speech, you wove an argument in support of your disliked position, you risked bringing about an unwanted consequence, and you were then given an attitude scale. Perhaps there was no opportunity for your compassionate memories to become accessible in memory.
Avoidance of Relevant Affirmations
Joshua Aronson, Hart Blanton, and I decided to assess this possibility (J. Aronson, Blanton, & Cooper, 1995). It was our plan to place people in a dissonant situation, make self-affirming cognitions readily accessible, and examine their impact on the dissonance process. It was not at all clear to us that making cognitions accessible that were directly related to the self-concept dimension threatened by the counterattitudinal advocacy would make people more comfortable. Although directly self-affirming on the threatened dimension, such cognitions might make the behavioral outcome seem worse and increase, rather than decrease, the uncomfortable arousal state of dissonance. Thus, such cognitions might be avoided rather than sought and not lower the need for dissonance-reducing attitude change.
In our study, we had volunteers participate in an experiment described as a study in personality structure. Participants were seated by a computer and were asked to respond on-line to the items generated on the screen. When they had completed the inventory, they were told that it would take approximately 15 min for the computer to analyze their responses and present a personality profile.
The study used a two-experimenter procedure, with participants being ushered into the office of a second experimenter while waiting for the computer to analyze the personality results. The second experimenter explained that he was working with the university on research about handicapped facilities on campus. He explained that the university was considering curtailing some of its services for the handicapped, and that the policymakers wanted to study students’ opinions. In keeping with procedures that have often been used in counterattitudinal advocacy research, the experimenter explained that one of the best ways to obtain people’s thoughts on both sides of an issue was to have people write strong and forceful essays favoring one of the two sides. He explained that he had plenty of essays opposing any curtailing of services for the handicapped, and that what he needed now were essays taking the strong and forceful position that services should be reduced. The first independent variable was the magnitude of the choice that participants were given to write the essay. In the high-choice conditions, participants were told that the choice to write this essay was completely their own, whereas in the low-choice conditions, they were merely told to write the antihandicapped services essay.
When the essay was completed and before any assessment of the participant’s attitudes, the participant was sent back to the office of the first experimenter to complete the personality structure research. The timing of the attitude assessment, relative to the participant’s opportunity to learn how compassionate he or she was, served as the second independent variable in the study. In the attitude-first condition, the second experimenter immediately knocked on the first experimenter’s door. Embarrassed and somewhat out of breath, he explained that he had forgotten to give the participant a questionnaire that the university needed. This questionnaire contained the crucial measurement of the participant’s opinion about reducing handicapped funding. In the attitude-second condition, the embarrassed first experimenter knocked on the door after the participant’s opportunity to receive information on how highly compassionate he or she was.
The first experimenter explained to all participants that the feedback from the personality profile was ready and there was both good news and bad news. The good news was that the computer had identified a large number of factors on which the participant scored greater than the median. The computer printed a paragraph describing the participant’s personality profile on all items in which he or she scored above the median. The bad news was that so many items had been identified, there was not time to read all of them. Thus, the participant was told he or she would have to choose which of 10 personality dimensions to read. Would the participants seek or avoid the directly affirming paragraphs that identified him or her as high in compassion? The participants’ choice of which paragraphs to read served as a second important dependent measure.
First, let us look at the attitude measure. Recall that in the attitude-first condition, participants had their attitudes assessed before receiving any feedback that could have served as self-affirmation. Both self-affirmation theory and dissonance theory would predict greater opinion change for high-choice than for low-choice participants (i.e., the typical induced-compliance effect found in many studies). However, in the attitude-second condition, participants had already had an opportunity to read about their good qualities, including—if they so desired—direct confirmation of their goodness in the area of compassion. This manipulation made positive self-information both available and accessible and should have provided ample opportunity for self-affirmation. Nonetheless, Figure 9.1 shows that reading about one’s good personality traits had absolutely no effect on attitudes. In the attitude-second condition, just like the attitude-first condition, high-choice participants expressed attitudes that were significantly more in favor of reducing handicapped funding than did participants in an attitude-only control condition in which attitudes were assessed without any experimental manipulations. Low-choice participants, however, did not differ from the control. The analysis of the data revealed a main effect for choice and significant simple effects for choice within the attitude-first and attitude-second conditions. Simply put, the opportunity to self-affirm had no impact on reducing the need to change attitudes following freely chosen, counterattitudinal advocacy.
FIGURE 9.1. Attitudes Toward Handicapped Funding After Counterattitudinal Behavior
Low numbers indicate greater attitude change. Adapted from “From Dissonance to Disidentification: Selectivity in the Self-Affirmation Process,” by J. Aronson, H. Blanton, and J. Cooper, 1995, Journal of Personality and Social Psychology, 68, p. 989. Copyright 1995 by the American Psychological Association.
We now address a related question: Did participants even want to see how compassionate they were? It follows from the basic assumptions of self-affirmation that people who have had their compassionate selves compromised by their counterattitudinal advocacy should jump at the opportunity to reinstate their compassion. The results, presented in Figure 9.2, showed that this did not happen. To the contrary, a main effect for choice showed that the more dissonance the participants had, the less they wanted to read about their compassion. And within the attitude-second condition, high-choice participants particularly wanted no part of learning about their compassion, showing the greatest avoidance of the compassion paragraphs of all the participants in the study.
FIGURE 9.2. Interest in Reading Affirming Information About Compassion as a Function of Choice and Affirmation Opportunity
Dependent measure is the proportion of paragraphs chosen about compassion relative to the total number of paragraphs chosen. Adapted from “From Dissonance to Disidentification: Selectivity in the Self-Affirmation Process,” by J. Aronson, H. Blanton, and J. Cooper, 1995, Journal of Personality and Social Psychology, 68, p. 990. Copyright 1995 by the American Psychological Association.
From this study, conducted at Princeton University, and a nearly identical replication run at Stanford University, we have learned a number of important facts: First, when in the throes of cognitive dissonance, people do not want to know about affirming, positive information relevant to the attribute that has been challenged by the potential consequences of their counterattitudinal behavior. Second, not only do participants not want such information, but its increased accessibility does not alter their need to reduce dissonance. Instead, participants continue to deal with the unwanted consequences of their behavior by changing their attitudes, presumably to render those consequences less aversive than they otherwise would have been.
Confronting People With Affirmation
The next question that my colleagues and I tried to answer was the effect of confronting people directly with self-affirming information (Blanton, Cooper, Skurnik, & Aronson, 1997). The previous study had shown that people who are experiencing dissonance avoid information that could directly affirm the valued personality trait that had been compromised by their counterattitudinal behavior. What would happen if people were presented with the information that showed that they were indeed compassionate, helpful people? Would making such praiseworthy information immediately salient help to reduce people’s discomfort, that is, would it reduce the need for attitude change?
Our prediction was that although a directly relevant affirmation might bolster a person’s self-esteem, it also could make salient a personal standard to which the behavioral outcome of an action could be compared. If people were forced to consider that they are helpful, compassionate people, that might cause the ideographic, personal standard to be brought to bear on the dissonant act.
Now, writing an essay against funding for the handicapped is aversive when considered against the normative standard and when considered against the personal standard of compassion. If compassionate people bring about effects that are antithetical to compassion, then the counterattitudinal behavior would be especially aversive. In summary, supplying a person with relevant, self-affirming information can lead to greater dissonance because the accessible personal standard makes the behavioral outcome seem even more noxious and aversive than when considered against the normative standard alone.
As in the previous study, Blanton et al. (1997) had participants volunteer for a study on personality structure. Participants were then asked by a second experimenter to write an essay arguing against funding for handicapped services at their university, under conditions of high or low choice. Participants returned to the first experimenter and read a paragraph that described an aspect of their personality that was identified by the computerized personality inventory. Some participants received positive, relevant feedback: They were told that they had scored highly in the trait of compassion. Other participants received positive, irrelevant feedback: They were told that they scored highly in the trait of creativity. A control group received no feedback at all.
A knock on the door brought the return of the out-of-breath, embarrassed second experimenter. As in the J. Aronson et al. (1995) procedure, he explained that he had forgotten to administer the questionnaire that the university wanted all participants to complete. This questionnaire included the major dependent measure: the assessment of attitudes toward handicapped funding.
The results showed strong evidence for our predictions. As can be seen in Figure 9.3, we found a main effect for choice and an interaction between choice and affirmation. In keeping with dissonance theory predictions, the results for the control (no-feedback) conditions showed a significant simple effect for choice: High-choice participants expressed attitudes that were more in line with the position of their essay than did low-choice participants. When participants read information that showed how positively compassionate they were (relevant-affirmation conditions), dissonance was increased. The same simple effect between high and low choice that had been found in the control condition was also found in the relevant-affirmation condition, but the effect was larger. Finally, the 2 × 3 factorial resulted in a significant interaction between choice and affirmation, aided by the lack of any difference between choice conditions when the affirming information was irrelevant to the topic of the essay. Thus, facing directly affirming information drives people further toward the need for dissonance reduction and increases, rather than decreases, dissonance-produced attitude change.
FIGURE 9.3. Attitudes Toward Proposing Relevant and Irrelevant Affirming Information
From “When Bad Things Happen to Good Feedback: Exacerbating the Need for Self-Justification With Self-Affirmations,” by H. Blanton, J. Cooper, I. Skurnik, and J. Aronson, 1997, Personality and Social Psychology Bulletin, 23, p. 689. Copyright 1997 by Sage. Reprinted with permission.
The Fuzzy (but Temporary) Good Feeling of Self-Affirmation
There is evidence in the Blanton et al. (1997) study that suggests that positive information people receive about themselves does seem to reduce the need for attitude change, provided that the information is not relevant to the self-attribute threatened by the counterattitudinal behavior. Note the lack of attitude change in the high-choice, irrelevant-feedback condition depicted in Figure 9.3. If direct self-affirmations result in increased dissonance, why does the indirect affirmation lead to less dissonance?
We think that the reason people show less dissonance motivation when receiving positive, but irrelevant, information is that they are literally distracted from the task at hand. They are distracted from the behavioral consequence by the fuzzy, good feeling of learning about a good attribute they possess. As we have argued, responsibly creating an aversive consequence creates arousal that is experienced as negative affect. When people add to that hedonic equation some positive news about themselves, their discomfort is reduced, and the exigent need to change their attitude is diminished.
However, such positive affect is fragile. We already know, from the results of the research presented here, that positive information will not lead to distracting, positive affect if it serves to make accessible a personal standard that has been directly violated by the attitude-discrepant behavior. We suspect that even irrelevant information (i.e., information that does not make the violated personal standard salient) will reduce negative affect only temporarily. It can fail: People can cease paying attention to it, or the information might be shown to be wrong. In such cases, we predict that the dissonance will return with a magnitude at least as great as that which existed before the positive, irrelevant information was introduced.
Galinsky, Stone, and Cooper (2000) conducted a study to examine the latter hypothesis. We allowed participants in a dissonance experiment to think of some of their positive self-values. Similar to previous affirmation experiments, we facilitated participants’ basking in this information and feeling good about its implications by giving them the AVL. Then, for some participants, we negated the information by providing information that questioned whether they indeed possessed these positive values. We predicted that the participants would no longer have their fuzzy, good feeling. Instead, they would need to reconfront their psychological discomfort, and the need to change their attitude would reemerge.
During the 1990s, there was a tradition at Princeton University that was enthusiastically supported by undergraduate students and opposed by the town police and university officials. It occurred on the day of the first snowfall of the season and was euphemistically known as the Nude Olympics. This now-defunct tradition provided a particularly good context to conduct our induced-compliance procedure. We asked (or told) students to make speeches arguing for an end to their treasured Nude Olympics. Then, in keeping with the procedure of Steele and Liu (1983), participants filled out the AVL, which gave them the opportunity to express themselves on valued dimensions of their personality.
Some participants immediately filled out an attitude survey that asked their positions on a number of campus issues, including the degree to which they supported the Nude Olympics. Other participants were given the opportunity to affirm the self by expressing their values on the AVL. Two other conditions, one that performed the dissonant behavior under conditions of high choice and the other under conditions of low choice, were randomly assigned to the disconfirmed condition. These participants received alleged feedback based on their responses to the AVL. They were shown graphs that placed their scores in the context of other university students. The graphs showed them that the positive values they thought characterized their personalities were, in fact, not characteristic of them at all. Essentially, their affirmations were negated and disconfirmed. Following the viewing of the graphs, participants in the affirmation and disconfirmed conditions filled out attitude scales and other measures.
The results of the attitude measure are depicted in Figure 9.4. We found that participants who were distracted by the positive information about their values did not manifest attitude change toward the Nude Olympics and differed significantly from the high-choice-no-feedback participants. However, when the participants learned that they had scored low on those treasured values that they thought characterized their personalities, their attitudes reverted to those expressed by the high-choice, no-feedback participants. In other words, the relief from needing to deal with the aversive consequences of their behavior was short-lived. As soon as the positive information was disconfirmed, the need to deal with the reality of the aversive event returned and so did dissonance-produced attitude change.
FIGURE 9.4. Attitudes Toward the Nude Olympics After Affirmation and Negation (Disconfirmation)
From “The Reinstatement of Dissonance and Psychological Discomfort Following Failed Affirmations,” by A. D. Galinsky, J. Stone, and J. Cooper, 2000, European Journal of Social Psychology, 30, p. 140. Copyright 2000 by John Wiley & Sons. Reprinted with permission.
Low-choice disconfirmed participants did not show justification of their dissonant behavior. This result indicates that it was not negative feedback per se that caused attitude change. Rather, discrediting participants’ values produced attitude change only for those participants who initially experienced dissonance and who had attempted to deal with the situation by focusing on their positive values.
In a separate experiment, using a very similar procedure, we examined the effect of affirming or negating positive, self-relevant information on the participants’ emotional experience. Using a measure developed by Elliot and Devine (1994), participants were asked to rate their affect after producing a counterattitudinal speech, after self-affirming on the AVL, or after receiving information that negated their affirmation, depending on the experimental condition to which they had been assigned. They were asked to rate their psychological discomfort by indicating the degree to which they felt uncomfortable, uneasy, and bothered.
In keeping with the attitude data presented in Figure 9.4, participants in the high-choice, no-feedback condition who did not have the opportunity to think about their favorite values showed a high degree of psychological discomfort, as dissonance theory predicts. On the other hand, participants who contemplated their positive values expressed a very low level of psychological discomfort, as predicted by self-affirmation. However, when the positive thoughts were negated by information showing that participants did not possess a high degree of those values, the psychological discomfort reemerged at a level similar to that of the high-choice, no-feedback participants. As in the previous experiment, low-choice, negative-feedback participants did not experience discomfort. This demonstrates that the negation information per se did not have a negative impact on the participants’ discomfort level. Rather, writing a counterattitudinal essay under high-choice conditions appears to be necessary to invoke the feeling of discomfort, a feeling state akin to Festinger’s (1957) original conceptualization of dissonance. Writing under low choice produced no psychological discomfort, even when it was accompanied by information showing people that the degree to which they held certain treasured values was not very high. Allowing people to bask in the belief that they did have treasured values reduced psychological discomfort, but the relief was fragile and fleeting. Discomfort reemerged in full force when the participants’ belief in their values was negated by the computerized feedback.
Putting the Pieces Together
In summary, the current portion of my argument is that dissonance reduction is not a strategy to make the self feel better or affirmed. Rather, it is the result of an unpleasant arousal state and is designed to render the consequence of a behavior nonaversive. Our data have shown that (a) people do not wish to find out about positive features of the self that have been threatened by attitude-discrepant behaviors, even though such information could directly repair the integrity of the compromised self; (b) people who are induced to confront affirming information directly manifest an increased need to reduce dissonance through attitude change; and (c) people who receive positive information that is not related to the compromised part of the self may show less of a need for attitude change after counterattitudinal behavior, but such reduction is temporary and fragile, with attitude change re-emerging if the affirmations fail. I am not arguing that people do not have an independent motive to feel good about themselves or to direct their actions toward restoring self-integrity. However, I am arguing that in situations in which people responsibly produce unwanted consequences, they must deal directly with the cognitive representation of those consequences—usually by changing their attitudes. Attempts to deal with their compromised self-integrity must wait until they have dealt with the unwanted consequences of their behavior.
SPECULATIONS ON A NEW ROLE FOR THE SELF IN DISSONANCE
Earlier, I indicated that there was no necessary role played by the self in dissonance reduction. That statement is now qualified to mean that dissonance is not confined to people with a particular self-view or to people with a particular level of self-esteem or to people with a particular personality trait. Much of the data presented in this chapter has focused on whether dissonance is merely one of many possible strategies to make a shattered self feel better and concluded that that was not a proper role for the self in dissonance theory.
However, the self is very much involved in dissonance if one takes a somewhat broader perspective (Stone & Cooper, 2001). First, the self plays a role as a standard against which a consequence can be considered wanted or unwanted. When self-standards are made highly accessible by environmental events or, in what I suspect are rare circumstances, by chronic accessibility, then they supersede normative standards in determining the aversiveness of a behavioral outcome (Stone & Cooper, 2003). Second, and somewhat more speculatively, the self may be intimately connected to the ontogeny of dissonance.
In the original new look paper (Cooper & Fazio, 1984) and a subsequent empirical study (Cooper, 1998), Fazio and I suggested that dissonance developed as a learned drive. Think about how children may learn to anticipate negative events in their lives. They learn that certain behaviors are followed by punishments and threats. Soon, children learn to anticipate the connection between behaviors and negative outcomes, and they avoid behaving in ways that bring such outcomes. Sullivan (1953), in his social psychiatric theory of personality development, discussed the creation of the self-system as a way of bringing about security while avoiding anxiety. The key to the system is that the self develops as a complex system of cognitions and behaviors, all designed to anticipate and cope with anxiety-producing reactions from people in the environment. It seems reasonable that one such negative, painful reaction from the environment occurs when a child brings about a consensually agreed, negative outcome. Hurting one’s mother, or a baby sister, or even knocking over a lamp may qualify as normatively defined negative events. Children thus learn to anticipate that such events lead to profound negative responses and are to be avoided. So an uncomfortable emotional reaction may develop at any hint or anticipation of responsibly bringing about an aversive event. If children do bring about an aversive event, then they need to develop ways to cope. One way might be to deny responsibility; if that fails, they may need to change what was once aversive into something nonaversive. What may develop as a response to the anxiety reactions and sanctions of significant people in the environment may eventually develop its own autonomy and become the tension state known as dissonance.
The gist of the analysis is to speculate that the development of ways to cope with producing aversive behavioral outcomes may be intimately related to the development of the entire self-system. Avoiding dissonance situations is learned as part of the self-system, as are coping mechanisms, should unwanted, aversive events occur. Dealing with significant others in the environment helps us form our self-systems, and, part and parcel of that development, is the development of cognitive dissonance.
Dissonance theory, throughout its long history, has been marked by extensions and applications into new and uncharted territory. Examining the role of the self in dissonance theory may provide another new opportunity. Although I have taken issue with the precise way that theories of self-consistency and self-affirmation view the involvement of self in dissonance, examining dissonance within the context of the self-system may prove a fertile area for future theory and research.
REFERENCES
Allport, G. W., Vernon, P. E., & Lindzey, G. (1960). Allport—Vernon—Lindzey Study of Values. Chicago, IL: Riverside.
Aronson, E. (1968). Dissonance theory: Progress and problems. In R. Abelson, E. Aronson, W. McGuire, T. Newcomb, M. Rosenberg, & P. Tannenbaum (Eds.), Theories of cognitive consistency: A sourcebook (pp. 5—27). Chicago, IL: Rand McNally.
Aronson, E. (1992). The return of the repressed: Dissonance theory makes a comeback. Psychological Inquiry, 3, 303—311. http://dx.doi.org/10.1207/s15327965pli0304_1
Aronson, J., Blanton, H., & Cooper, J. (1995). From dissonance to disidentification: Selectivity in the self-affirmation process. Journal of Personality and Social Psychology, 68, 986—996. http://dx.doi.org/10.1037/0022-3514.68.6.986
Blackman, S. F., Keller, K. T., & Cooper, J. (2016). Egocentrism and vicarious dissonance. Journal of Experimental Social Psychology, 62, 1—6. http://dx.doi.org/10.1016/j.jesp.2015.09.001
Blanton, H., Cooper, J., Skurnik, I., & Aronson, J. (1997). When bad things happen to good feedback: Exacerbating the need for self-justification with self-affirmations. Personality and Social Psychology Bulletin, 23, 684—692. http://dx.doi.org/10.1177/0146167297237002
Brehm, J. W. (1956). Postdecision changes in the desirability of alternatives. Journal of Abnormal and Social Psychology, 52, 384—389. http://dx.doi.org/10.1037/h0041006
Cooper, J. (1998). Unlearning cognitive dissonance: Toward an understanding of the development of dissonance. Journal of Experimental Social Psychology, 34, 562—575. http://dx.doi.org/10.1006/jesp.1998.1365
Cooper, J. (2007). Cognitive dissonance: Fifty years of a classic theory. Thousand Oaks, CA: Sage.
Cooper, J., & Fazio, R. H. (1984). A new look at dissonance theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 17, pp. 229—266). New York, NY: Academic Press.
Cooper, J., & Scalise, C. J. (1974). Dissonance produced by deviations from life styles: The interaction of Jungian typology and conformity. Journal of Personality and Social Psychology, 29, 566—571. http://dx.doi.org/10.1037/h0036217
Cooper, J., & Worchel, S. (1970). Role of undesired consequences in arousing cognitive dissonance. Journal of Personality and Social Psychology, 16, 199—206. http://dx.doi.org/10.1037/h0029830
Cooper, J., Zanna, M. P., & Goethals, G. R. (1974). Mistreatment of an esteemed other as a consequence affecting dissonance reduction. Journal of Experimental Social Psychology, 10, 224—233. http://dx.doi.org/10.1016/0022-1031(74)90069-9
Elliot, A., & Devine, P. (1994). On the motivational nature of cognitive dissonance: Dissonance as psychological discomfort. Journal of Personality and Social Psychology, 67, 382—394. http://dx.doi.org/10.1037/0022-3514.67.3.382
Festinger, L. (1957). A theory of cognitive dissonance. Palo Alto, CA: Stanford University Press.
Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58, 203—210. http://dx.doi.org/10.1037/h0041593
Galinsky, A. D., Stone, J., & Cooper, J. (2000). The reinstatement of dissonance and psychological discomfort following failed affirmations. European Journal of Social Psychology, 30, 123—147. http://dx.doi.org/10.1002/(SICI)1099-0992(200001/02)30:1<123::AID-EJSP981>3.0.CO;2-T
Hoyt, M. F., Henley, M. D., & Collins, B. E. (1972). Studies in forced compliance: Confluence of choice and consequence on attitude change. Journal of Personality and Social Psychology, 23, 205—210. http://dx.doi.org/10.1037/h0033034
Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480—498. http://dx.doi.org/10.1037/0033-2909.108.3.480
Kunda, Z. (1999). Social cognition: Making sense of people. Boston, MA: MIT press.
Norton, M. I., Monin, B., Cooper, J., & Hogg, M. A. (2003). Vicarious dissonance: Attitude change from the inconsistency of others. Journal of Personality and Social Psychology, 85, 47—62. http://dx.doi.org/10.1037/0022-3514.85.1.47
Scher, S. J., & Cooper, J. (1989). Motivational basis of dissonance: The singular role of behavioral consequences. Journal of Personality and Social Psychology, 56, 899—906. http://dx.doi.org/10.1037/0022-3514.56.6.899
Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 261—302). Hillsdale, NJ: Erlbaum. http://dx.doi.org/10.1016/S0065-2601(08)60229-4
Steele, C. M., & Liu, T. J. (1983). Dissonance processes as self-affirmation. Journal of Personality and Social Psychology, 45, 5—19. http://dx.doi.org/10.1037/0022-3514.45.1.5
Steele, C. M., Spencer, S. J., & Lynch, M. (1993). Self-image resilience and dissonance: The role of affirmational resources. Journal of Personality and Social Psychology, 64, 885—896. http://dx.doi.org/10.1037/0022-3514.64.6.885
Stone, J., & Cooper, J. (2001). A self-standards model of cognitive dissonance. Journal of Experimental Social Psychology, 37, 228—243. http://dx.doi.org/10.1006/jesp.2000.1446
Stone, J., & Cooper, J. (2003). The effect of self-attribute relevance on how self-esteem moderates attitude change in dissonance processes. Journal of Experimental Social Psychology, 39, 508—515. http://dx.doi.org/10.1016/S0022-1031(03)00018-0
Sullivan, H. S. (1953). Interpersonal theory of psychiatry. New York, NY: Norton.
I gratefully acknowledge the help of Adam Galinsky in the preparation of an earlier draft of this chapter.